Network Intrusion Detection Using Convolutional Neural Network

Network intrusion detection has become essential for maintaining the security and reliability of modern digital infrastructures. This study presents a deep learning–based intrusion detection system that leverages a Convolutional Neural Network (CNN) trained on the NSL-KDD dataset to classify network traffic into normal and multiple attack categories. The proposed model achieves a test accuracy of 82.24%, demonstrating its effectiveness in identifying diverse intrusion patterns. To enable practical real-time deployment, the system is integrated into an interactive Streamlit-based web application that allows users to submit input features, visualize attack distributions, and observe model predictions dynamically. This work highlights the potential of CNNs for intrusion detection and provides an accessible interface for further research, evaluation, and real-world experimentation in cybersecurity analytics. Key Terms: Intrusion Detection Systems (IDS), Convolutional Neural Networks (CNN), Deep Learning, Network Security, Cybersecurity Analytics, NSL-KDD Dataset, Anomaly Detection, Traffic Classification, Machine Learning Models, Real-Time Detection, Threat Identification, Attack Pattern Recognition, Data Preprocessing, Supervised Learning.